AI investment mania has altered the VC market as tech titans like Microsoft and Amazon invest billions into AI research projects.

There’s more than meets the eye with regard to this subject matter! Here at CARZONI Amazon Web Services CEO Adam Selipsky speaks with Anthropic’s Dario Amodei during AWS Re:Invent 2023 at The Venetian Las Vegas on Nov 28, 2023 in Las Vegas. Noah Berger | Getty Images Venture capitalists are facing tough circumstances three years into an otherwise dormant initial public offering cycle; their private markets are filled with highly valued AI startups described as generational firms vying for investment dollars from venture capitalists. Venture firms looking for exits won’t see relief anytime soon from artificial intelligence (AI), since unlike prior tech booms VCs don’t currently hold such prominence within it. Microsoft, Amazon, Alphabet and Nvidia — four giant companies in the industry — have committed billions to fuel the development of capital intensive companies like OpenAI, Anthropic, Scale AI and CoreWeave. As these institutions contribute billions towards funding this revolution of artificial intelligence innovation, their vast financial reserves don’t apply as usual when considering going public with these projects. Although they might show profits sooner, these startups don’t meet the profitability metrics required of public investors before taking the leap into investing with them.Tech giants boast much more than money. They offer tangible incentives like cloud credits and business partnerships – benefits which VCs cannot match. Melissa Incera, an analyst from S&P Global Market Intelligence told CNBC: “AI startups that we speak to are fundraising at healthy valuations without difficulty. Many still report too much investor interest at present.” “Venture investors face an unprecedented market distortion that shows no signs of abating.” According to an August 29 report from PitchBook, U.S. venture capital exit value this year may reach $98 billion, down 86% since 2021; venture-backed IPOs should see their lowest performance level since 2016. Traditional venture capital firms (VCs) are diving headlong into AI investments, but tend to favor investments higher up the stack: such as seed capital for startup apps requiring far less capital than infrastructure-backed AI solutions such as generative AI. So far in 2024 investors have pledged $26.68 billion into 498 generative AI deals including strategic investors’ contributions according to PitchBook. As 2023 showed us, when AI companies raised $25.9 billion total–an increase of more than 200% over 2022–Forge Global has noted an exponential surge in AI fundraising from just 12% last year to 27% so far this year. Data shows that AI companies had 140% larger rounds this year compared with last, while non-AI firms experienced only 10% growth versus last. According to Chip Hazard of early stage firm Flybridge Capital Partners, investment dollars are shifting up the stack and that “enduring companies will be formed at the application layer”. “That will take time.” Startup investors continue to feel the effects of market turmoil since early 2022 when rising inflation caused the Federal Reserve to raise interest rates, forcing investors into more prudent assets that offered greater returns – such as Nvidia chips used for training most AI models as well as megacap stocks such as Microsoft Meta and Amazon whose shares had surged back up the charts. Tech stocks, led by Nvidia which dominates AI training as well as Microsoft, Meta and Amazon have since rebounded substantially, propelling tech stocks forward once again. In July, the Nasdaq set a record before declining back slightly since. Unfortunately, few initial public offerings and pricey acquisitions have taken place so venture firms have seen few returns for their limited partners. PitchBook stated in its August report: “Managers have found it challenging to raise additional funds without providing their investors with returns,” and this challenge is compounded by more liquid, lower-risk investments having attractive yields due to high interest rates. Cerebras appears to be close to going public; founded in 2016 and supported by traditional venture capital funds Benchmark Capital and Foundation Capital among others. Cerebras never attained the lofty valuations seen among AI model developers or infrastructure players; its peak valuation in 2021 stood at $4 billion before markets experienced their sharp downward turn. Cerebras announced in late July that it had filed confidentially its initial public offering paperwork with SEC for approval. The company still has not submitted its public prospectus. Cerebras did not respond.When it comes to foundational model companies, their quickly attained valuations put them in an entirely separate league from VCs, said Jeremiah Owyang of Blitzscaling Ventures.According to Owyang it can be “very challenging for VCs right now to guarantee any exits due to market conditions; early-stage investors may see returns between 7-12 years on their newer bets. Menlo Ventures and Inovia Capital take an alternate approach in AI; Menlo revealed its plan in January to raise an “special purpose vehicle”, named Menlo Inflection AI Partners as part of a $750 million funding round for Anthropic that valued it at more than $18 billion. At Anthropic’s launch in 2021, Amazon was its primary investor to compete against Microsoft – who have invested billions into OpenAI and are reported as part of an upcoming funding round valuing ChatGPT creator at over $100 billion. Menlo had also previously made investments at about 4.1 billion valuation. Menlo had to look elsewhere in order to raise more capital at an increased price, outside its main $1.35 billion fund that closed last year. Menlo raised an SPV, whereby investors commit money directly towards one investment rather than investing across an array of companies. Menlo filed to raise $500 million through an SPV in April; Cohere, which offers enterprise AI capabilities through its AI platforms Cohero and Cohere AI for Business Solutions announced a $500 million funding round comprised of investors including AMD, Salesforce.Oracle,Nvidia that valued it at more than double what was seen last year – more than tripling its valuation since 2016. Cohere confirmed to CNBC that part of this funding came through SPVs for future fundraising activities as well. Menlo Ventures of Menlo Park organized Inovia’s SPV with Shopify CEO Tobias Lutke among those participating. Unfortunately, neither Menlo or Inovia responded to our inquiries for comments or responses regarding SPVs that allow multiple investors to pool capital together into one hot company. Some investment banks have also put together such vehicles in order to facilitate multiple investors pooling together their funds into promising startups. JPMorgan Chase recently told CNBC that clients of its Morgan Private Ventures unit “have access to several leading AI investments.” For investors to realize any return from these investments, however, an initial public offering must occur at some point; otherwise it becomes practically impossible for tech giants to undertake substantial acquisitions without first going public first. Companies such as Microsoft, Alphabet, Amazon and Nvidia can be patient investors; between them they hold over $280 billion of cash and marketable securities on their balance sheets.The IPO pipeline will continue to build’One path to liquidity is secondary market investing – selling shares to another investor directly – though Elon Musk’s SpaceX (recently valued by employee tender offer at over $200 billion) also enabled secondary transactions through secondary shares transactions. At least that may be in store for investors in xAI, Musk’s 18-month old AI startup that already stands at $24 billion following an enormous $6 billion round raised last May. But SpaceX stands apart. Secondary transactions have traditionally been seen as an avenue for founders and early investors to cash out a portion of their stake in an overvalued startup company, rather than as an opportunity for venture capitalists to make returns on investments in such companies. For this to work they require initial public offerings (IPOs). SpaceX’s Polaris Dawn Falcon 9 rocket sits atop Launch Complex 39A of NASA’s Kennedy Space Center on August 26, 2024 near Cape Canaveral in Florida and was taken for launch -Joe Raedle/GettyImages Michael Harris, global head of capital markets at the New York Stock Exchange told CNBC recently that his exchange is currently engaging with “a number of AI-focused companies”, adding: As industry progresses we would anticipate this pipeline continuing to grow.” Only a select few AI companies have ventured onto the public markets this year, including Astera Labs’ debut on Nasdaq as data center connectivity vendor to cloud and AI infrastructure companies in March. Tempo AI, backed by Google and going public last June, was valued at approximately $6.5 billion as it went on the market for trading on its debut day of trading. Stock of this artificial intelligence (AI) company has since gained around 50 percent since its debut, giving a valuation of $8.6 billion. Unfortunately, though, no other prominent AI companies are discussing going public anytime soon. “Unless there’s been a substantial shift in market sentiment, I find it hard to comprehend why AI startups would put themselves under such scrutiny when their growth can continue in private at such favorable terms,” noted S&P’s Incera. “Coming public would only add pressure for them to show returns or reduce expenditure, something which may not be realistic at this stage in their maturity curve,” according to she said. Most venture investors remain bullish on generative AI’s potential to deliver big dividends at the application layer level. Amazon, Google and Facebook all had web apps built upon internet infrastructure that revolutionized tech cycles over time. Uber, Airbnb and Snap are among the many valuable apps created on smartphone platforms, which John-David Lovelock of Gartner sees an opportunity for. He believes there may be great promise for Generative AI within enterprise settings. Yet by 2024, only one percent of software spending will come from companies purchasing GenAI-related products, according to Lovelock. While money may be spent on particular GenAI tools and applications that exist today, Lovelock says its widespread implementation within enterprise software catalogue has yet to take place. Don’t miss these insights from CNBC PRO!